A STUDY IN CAECINOLOGY. 31 
together in the same connective-tissue shealh posteriorly, leaving onlj^ a 
small circular j^assage between them anteriorly. The connectives between 
the next succeeding ganglion-pairs are pretty widely divaricated to admit 
the passage of the sternal artery, and the last two thohicic ganglion-pairs are 
so intimately fused that the separate ganglia cannot be distinguished. The 
abdominal ganglia are precisely like those of Ranina, but somewhat larger in 
proportion to the rest of the ventral chain. In Notopus dorsvpes and 
Notosceles chimmonis the ventral nervous chain presents much the same 
characters, but my examples of these species had lain for many years in spirit 
and were too rolten to admit of exact study. In Notopus the ganglion-pair 
of the chelipeds is separated from the suboesophageal ganglion mass by an 
interval somewhat shorter than in Lyreidus, but the connectives joining them 
are distinctly paired and are not included in a common connective-tissue 
sheath. 
The above descriptions, with the figures illustrating them, show that the 
thoraco-abdoininal nerve system of the Raninida3 approximates more nearly 
to the Macruran type than does that of any other Brachyuran — indeed, more 
nearly than any Anomnran except Hippa, in which genus, as I have 
ascertained by dissection, the thoracic ganglia are spaced fairly evenly along 
the longitudinal line and are united by distinctly paired objectives, except 
the last two which are fused together. The abdominal chain also in Hippa 
consists of at least five fairly distant ganglion-pairs united by distinctly 
paired connectives, but I have not 3 et finished my observations on this genus, 
and must postpone the description of its nervous system to another paper. 
As compared with the Decapoda, of which the nervous systems have lieen 
accurately described and figured, the abdominal ganglia of the Uaninidfe 
most nearlj' resemble the condition found in Dromia vulgaris, as described 
and figured by Bouvier (22). I have myself verified the accuracy of this 
(lesciiption. But in Dromia the thoracic ganglia, though distinct, are 
closely api^roximated in longitudinal series, and form a sort of oval beaded 
ring round the perforation for the passage of the sternal artery, as is the 
case in Calappa among the Oxystomes, in Corystes, and generally in the 
Catametopan section of the Brachygnatha. In respect of the longitudinal 
concentration of the thoracic ganglia, the Raninidre exhibit a much more 
primitive condition than Dromia. Similarly, the extended thoracic chain of 
the Raninidse is in a much more primitive condition than are the approxi- 
mated though distinct thoracic ganglia of Galathea and Porcellaiia among 
the Anomura, though these are more primitive as regards the abdominal 
chain. The same may be said in respect oi Eupagurus,ii\ which the thoracic 
ganglia are in some respects more completely fused together than in the 
Galatheidse. If we carry on the comparison to the Macruran typos, the 
Haninidse cannot be brought into close relation with the Lobsters, and still 
